Modelling, Parameter Estimation and Control of Biodiesel Microreactors

Abstract

Globally, there is a growing need to reduce the overall carbon emission from various human activities. As a result, there are ongoing innovative ideas aimed at developing alternative energy sources which are less toxic to the environment. One such is the production of biodiesel in microreactors. This thesis captures a holistic outlook of the modelling, parameter estimation and control of biodiesel microreactors. The work also extends to practical applications and experiments from a microreactor rig based at Reutlingen University, Germany. The general modelling concepts presented here are applied to a microreactor built with accompanying peripheral sensing and actuation devices. These devices include a near-infrared (NIR) sensor for capturing online measurements, high-performance liquid chromatography (HPLC) pumps for a precise flow of reactants, pressure and temperature sensors and a carefully designed reaction chamber. Candidate mathematical models for the control of generic biodiesel microreactors are proposed. These models are intrinsically nonlinear with combinations of convection, diffusion and stoichiometric effects. The proposed mod- els are in nonlinear state space forms with some unmeasured system states. This poses difficulties in estimating such model parameters. A Sequential Monte Carlo method, particularly particle filtering and smoothing is proposed to circumvent this problem. This method is applied to estimate the parameters of one of the proposed nonlinear models. The estimated model is validated with independent experimental data sets. The experimental results reinforce the suitability of Sequential Monte Carlo approximation for estimating parameters of systems with limited identification data and unmeasured states. Dual Youla parameterisation is also proposed as a method for developing an empirical model of the biodiesel microreactor with a stability guarantee. With this parameterisation and due to the close relationship between the Youla parameterisation and the Internal Model Control (IMC) architecture, a modified nonlinear Internal Model Control (NLIMC) framework is further proposed for the control of biodiesel microreactors. The modified NLIMC outperforms the basic linear and nonlinear IMC mainly when input saturation is present in the system.

Date of Award	1 Aug 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Arthur Garforth (Supervisor) & William Heath (Supervisor)